SIRT6

Abstract

Sirtuins are the mammalian homologs of the yeast histone deacetylase Sir2. In recent years, an ever-expanding picture has emerged indicating that these proteins (SIRT1-7) play broad functions in cellular stress resistance, genomic stability, energy metabolism, aging and tumorigenesis.¹ Among members of this family, SIRT6 appears to have particular significance in regulating metabolism, DNA repair and lifespan.^2–4 In this context, new research from our lab has established SIRT6 as a key regulator of glucose homeostasis.⁵ In this Point of View article, we will first highlight our recent findings, and then provide an indepth discussion of their implications in cancer and aging.

SIRT6 Regulates Glucose Metabolism

The first clue that SIRT6 might play a role in glucose metabolism came from the phenotype observed in SIRT6 deficient mice. Although relatively normal at birth, they rapidly develop a variety of degenerative conditions, including complete loss of subcutaneous fat, lymphopenia, osteopenia and lordokyphosis. The most striking phenotype, however, is acute onset of hypoglycemia, which eventually kills the mice before one month of age.⁴ This phenotype is not triggered by hyperinsulinemia. Rather, the pancreatic islets in these mice are normal and blood insulin levels are actually lower than those of wild-type mice, indicating that SIRT6 deficiency induced insulin hypersensitivity, possibly an adaptive response to the low glucose levels. In contrast, both skeletal muscle and brown adipose tissue exhibit a remarkable increase in glucose uptake in these animals, likely explaining the glucose depletion in the blood.

This enhancement in glucose uptake was recapitulated in cultured SIRT6 deficient Embryonic Stem (ES) cells and Mouse Embryonic Fibroblast (MEF) cells, indicating that SIRT6 is a cellautonomous regulator of glucose homeostasis. Furthermore, we showed that overexpression of the glucose transporter GLUT1 accounted for the elevated glucose intake in these cells.⁵ We next wished to address whether SIRT6 deficiency is associated with aberrant glucose utilization in addition to this alteration in transport. Normally, glucose is converted into pyruvate through a series of glycolytic reactions. Pyruvate is then either converted to lactate by the enzyme lactate dehydrogenase (LDH) or is shuttled into mitochondria as a substrate in the Tricarboxylic Acid Cycle (TCA) leading to eventual metabolism through oxidative phosphorylation (Fig. 1). The TCA/OxPhos route is the most efficient way to produce ATP to fuel all the cellular biological activities (36 molecules of ATP per molecule of glucose). Cells favor this pathway under normal oxygen and nutrient conditions. On a per mole basis, the LDH route is much less effective in producing ATP (2 molecules of ATP), yet it minimizes the requirement for oxygen availability and the TCA intermediates can be used for anabolic reactions critical for cell survival (e.g., fatty acid, amino acid and nucleotide biosynthesis).⁶ Thus, under conditions of low nutrient availability or other stress conditions (e.g., hypoxia), cells switch towards lactate production as an adaptive survival response.

Figure 1.

Schematic diagram of glycolysis. See text for details.

In SIRT6 deficient cells grown in nutrient replete conditions, there is a marked switch in glucose metabolism that favors lactate glycolysis: glucose uptake and lactate production increase, whereas oxygen consumption and ATP production decrease.⁵ Hence, these cells behave as though they are experiencing glucose shortage or nutrient stress, so that metabolism is converted from “growth mode” to “survival mode”, suggesting that SIRT6 plays a critical role in sensing nutrient levels in the environment.

What is the molecular basis for this metabolic change? First, multiple key glycolytic genes show increased expression patterns. For instance, Lactate Dehydrogenase (Ldh) is upregulated, explaining the increased lactate production. Other genes that show increased expression include triose phosphate isomerase (Tpi), aldolase, Glut1, and the rate-limiting glycolytic enzyme phosphofructokinase-1 (Pfk-1). Notably, Pyruvate Dehydrogenase Kinase genes Pdk1 and Pdk4 also show higher expression. PDK phosphorylates and inactivates pyruvate dehydrogenase (PDH), a rate-limiting enzyme that converts pyruvate to Acetyl-CoA to fuel the TCA cycle (Fig. 1). Therefore, increased expression of the pdk genes inhibits mitochondrial respiration by preventing pyruvate from entering the Krebs cycle. Overall, SIRT6 deficiency appears to simultaneously influence expression of genes affecting both forks in glucose utilization, enhancing its conversion to lactate and blocking its use in OxPhos.

How does SIRT6 regulate these genes? Chromatin Immunoprecipitation (ChIP) analysis of several glycolytic genes shows that SIRT6 directly binds to their promoter regions, and subsequently deacetylates histone H3K9—a mechanism previously linked to gene silencing.^5,7 Therefore, SIRT6 deficiency causes an increase in H3K9 acetylation in those promoters resulting in increased expression of these specific genes. Interestingly, in wild-type cells, RNA polymerase II (RNAPII) seems to be loaded onto the SIRT6-repressed promoters (as illustrated by Ldhb High Resolution ChIP analysis⁵), but remains stalled under normal nutrient conditions. As a consequence, minimal Ldhb RNA is generated. This represents a classical “poised gene” scenario where genes requiring rapid activation are engaged with paused RNAPII and are ready to be transcribed should environment changes call for it.⁸ In SIRT6 KO cells, however, this restriction is lifted, RNAPII moves along the DNA strand and expression of these genes is triggered. Exactly how SIRT6 imposes this restriction on RNAPII remains unclear. It would be interesting to determine which transcriptional elongation factors are affected by SIRT6, whether SIRT6 directly interacts with these factors, and to establish the specific role played by SIRT6-dependent deacetylation in their regulation. In particular, it is important to establish whether this repressive effect occurs via SIRT6-dependent deacetylation of the elongation factors or the H3K9 deacetylase activity affects their recruitment to the chromatin.

The identification of additional players provided more clarity in resolving this regulatory puzzle. Given that glucose metabolism is fundamental for cell survival, it is not surprising that it is kept under tight control. How does this newly discovered SIRT6 modulation fit into previously defined regulatory pathways? It turns out that SIRT6 accomplishes the job by interacting with another important glycolytic regulator, Hypoxia Inducible Transcription factor 1α (Hif1α). Hif1α is a key mediator in cellular adaptation to nutrient and oxygen stress. On one hand, it enhances glycolytic flux by upregulating expression of key glycolytic genes. On the other hand, Hif1α directly inhibits mitochondrial respiration by upregulating expression of the Pdk genes.^9,10 Overall, Hif1α appears to modulate multiple genes in order to activate glycolysis and at the same time repress mitochondrial respiration in a coordinated fashion. Hif1α abundance is controlled by a regulatory machinery that alters Hif1α protein stability in response to changes in environmental conditions. Under normoxia, Hif1α is hydroxylated in multiple prolyl residues by the prolyl-hydroxylase-domain (PHD) proteins. Following hydroxylation, Hif1α is recognized by the von-Hippel-Lindau (VHL) ubiquitin ligase, marking Hif1α for subsequent proteasome degradation. When O₂ is low, PHD is inactivated, thereby stabilizing Hif1α protein levels.¹¹ However, it is important to point out that even under normoxic and normoglycemic conditions, Hif1α regulates basal expression of its target genes,¹² suggesting that further mechanisms are in place to ensure that this stress response is tightly regulated under normal nutrient conditions.

Here is where SIRT6 comes in play. We noticed that many SIRT6 target genes are also targets of Hi1a (Fig. 1), indicating that the two may be functionally linked. Our studies revealed that SIRT6 can corepress Hif1α when tested on a luciferase reporter assay. Furthermore, these proteins physically interact under normal nutrient conditions, suggesting that in the presence of SIRT6, Hif1α transactivation of these promoters is inhibited.⁵ In addition, Hif1α protein levels increased in the absence of SIRT6, likely secondary to a positive feedback response in order to maintain this glycolytic switch. Finally, treating SIRT6 deficient cells with a Hif1α inhibitor¹³ or stably knocking down Hif1α in these cells not only rescues the glucose uptake phenotype, but also lowers expression of the cohort of glycolytic genes described above. Remarkably, injecting SIRT6 KO mice with the Hif1α inhibitor temporarily rescues the hypoglycemic phenotype, confirming that SIRT6 also exerts its effect in vivo through a Hif1α dependent mechanism. More importantly, it indicates that without SIRT6, Hif1α is ectopically activated even under normoxic/normoglycemic conditions, in line with the idea that SIRT6 deficiency mimics a nutrient-deficient environment.

Thus, a model of how SIRT6 regulates glucose metabolism emerges (Fig. 2): under normal nutrient conditions, SIRT6 binds to the promoters of glycolytic genes and keeps histone H3K9 acetylation levels low. At the same time, it co-represses Hif1α on these promoters, acting as a safeguard to avoid ectopic activation of Hif1α target genes. By doing so, SIRT6 directs glucose away from glycolysis and into the mitochondria for efficient ATP production. Under conditions of nutrient scarcity, however, SIRT6 is likely inactivated, causing Hif1α to become activated and glycolytic genes to be transcribed in order to direct this metabolic transition from mitochondria respiration to glycolysis. Therefore, SIRT6 seems to function as a nutrition and stress sensor and contribute to important cell fate decisions—e.g., survival vs. proliferation—by directing glucose away from glycolysis and into the mitochondria. Notably, a recent study using transgenic animals overexpressing SIRT6 shows that extra levels of SIRT6 causes enhanced glucose tolerance and increased glucose-stimulated insulin secretion under conditions of high-fat diet,¹⁴ further supporting a critical role for SIRT6 in glucose metabolism.

Figure 2.

Model of SIRT6 function. Under normal nutrient conditions, SIRT6 inhibits expression of glycolytic genes, functioning as a histone deacetylase to co-repress Hif1α. This maintains proper flux of glucose to the TCA cycle. Under nutrient stress conditions, SIRT6 is inactivated, allowing activation of Hif1α, recruitment of p300, acetylation of H3K9 at the promoters, and increased expression of glycolytic genes, causing increased glycolysis and reduced mitochondrial respiration.

However, several missing pieces remain to be fully explored. For example, how is SIRT6 inactivated under nutrient deficient or stress conditions? Is it removed from chromatin, or does it remain bound but somehow become enzymatically inactive? Another unanswered question is how SIRT6 co-represses Hif1α. Previous studies have shown that Hif1α activity depends on its ability to recruit the histone acetyl-transferase p300.^15,16 Therefore, one possibility could be that binding of SIRT6 to Hif1α (and subsequent H3K9 deacetylation on the glycolytic gene promoters) “competes out” p300, precluding activation of these genes. In this context, recent studies have shown that another sirtuin, SIRT1, can regulate expression and activity of both Hif1α and Hif2α.^17,18 Although these studies were performed in vitro, and the role for SIRT1 in glucose metabolism remains to be fully explored, they indicate that activity of Hif factors might be regulated by the coordinated action of multiple sirtuins, a possibility that will likely be explored in the future.

SIRT6: A Potential Tumor Suppressor Gene?

In his seminal studies in the 1920s, Otto Warburg showed that cancer cells undergo altered glycolysis that favors lactate production even in the presence of adequate oxygen.¹⁹ Termed the “Warburg Effect” or aerobic glycolysis, it is considered as one of the hallmarks of malignant cells.²⁰ As detailed above, loss of SIRT6 renders cells glycolytic production towards lactate production, even under aerobic conditions, suggesting that lack of SIRT6 might provide a growth advantage for tumor cells. Several lines of evidence support this notion. First, SIRT6 deficient ES cells are more resistant to apoptosis when exposed to hypoglycemia or hypoxia, another hallmark of cancer cells.^5,20 Second, Hif1α, a major mediator of aerobic glycolysis that is overexpressed in SIRT6 deficient cells, is found at elevated levels in many primary human cancers.²¹ Third, one of the SIRT6 interacting protein, GCIP, is a potential tumor suppressor that is downregulated in a number of cancers.²² Mice with germline SIRT6 deficiency die at an early age and hence the development of a SIRT6 conditional strain will be a valuable tool in addressing this question in vivo.

SIRT6 and Aging

The discovery that the yeast sirtuin homolog, Sir2, promotes longevity has prompted intensive studies in the aging field. It has now been shown that sirtuins participate in life-span regulation in a variety of additional species, including worms and flies.^1,23 In mammals, most of the attention has been focused on SIRT1, which modulates multiple regulators of aging and is possibly involved in the beneficial effects of caloric restriction.²⁴ Among the sirtuin knockout mice, however, SIRT6 deficient animals have the most striking phenotype with several features that can be considered to resemble “premature aging”, as described above. Recently, Chua and colleagues showed that SIRT6 can repress the NFκB signalingpathway, a pathway that has been linked to aging in many mammalian tissues and in stem cells.^2,25 Our studies now demonstrate a critical role for SIRT6 in glucose homeostasis. Notably, food intake restriction and balanced metabolism have long been established as fundamental to attaining a healthy lifespan.²⁶ In this context, increased glucose uptake shortens lifespan of C. elegans via transcriptional regulation,²⁷ and recent studies have shown that Hif1α can also modulate lifespan in these organisms.^28,29 Even though the effect of glucose intake and Hif1α in mammalian aging remains to be explored, the ability of SIRT6 to link glucose/oxygen availability to transcriptional outputs highlights the intriguing possibility that SIRT6, in coordination with Hif1α, might play a role in mammalian aging by modulating metabolic homeostasis.

Perspective

SIRT6 has emerged as a critical regulator of metabolism, likely influencing multiple biological processes, such as cell proliferation, energy and nutrient balance, and potentially contributing to metabolic diseases and tumorigenesis. Undoubtedly, we have just begun to scratch the surface in terms of elucidating the biological functions of SIRT6, as novel functions are likely to be identified for this fascinating chromatin factor. In this context, further research will help to determine whether modulating SIRT6 activity could provide therapeutic benefits under conditions of metabolic imbalance, as observed in diseases like obesity, diabetes and cancer. SIRT6 is a major player in energy homeostasis, and the stage is now set to fully discover the activities of this important chromatin protein in mammalian pathophysiology.